Risk stratification of in-hospital venous thromboembolism for urological cancers: a multicenter retrospective study

Background: The current predictive models for venous thromboembolism (VTE) have limitations in predicting VTE risk in patients with urological cancer. Objectives: To establish and validate a risk stratification model for in-hospital VTE in patients with urological cancer. Design: Retrospective, multicenter study. Methods: The clinical data of 735 patients with urological cancer in the Department of Urology at four hospitals in China between January 2019 and December 2024 were analyzed. VTE (n = 147) and non-VTE (n = 588) groups were formed based on inclusion and exclusion criteria to develop a predictive model for VTE in patients with urological cancer. Results: In this study, we developed a risk stratification model using a logistic regression based on variables selected by LASSO and constructed a nomogram to visualize the model. The areas under the receiver operating characteristic curve for the training, validation, and external validation cohorts were 0.933 (95% confidence interval (CI): 0.909–0.957), 0.900 (95% CI: 0.850–0.950), and 0.857 (95% CI: 0.776–0.938), respectively. The corresponding Brier scores for them were 0.070, 0.089, and 0.200. The calibration curve indicated good model performance. Decision curve analysis evaluated the clinical utility of the model and showed that it provided a higher net benefit than the “treat-all” and “treat-none” strategies across a wide range of threshold probabilities, supporting its potential clinical usefulness. Conclusion: A predictive model was developed to estimate the risk of in-hospital VTE in patients with urological cancer, enabling personalized assessment and guiding preventive strategies. Further studies are needed to better validate our model.